the future of everything: the science of prediction

BOOK REVIEW
David Orrell’s
THE FUTURE OF EVERYTHING:
THE SCIENCE OF PREDICTION
reviewed by Paul Fields
I
f you make forecasts, or if you use forecasts,
you should read David Orrell’s book, The
Future of Everything – an engaging, as well as
deeply insightful, discussion on the difficult task of
prediction.
Throughout history, people have tried to foretell the
future. In The Future of Everything, Orrell traces
the history of forecasting from ancient Greece to
today, outlining the many ways people have tried to
be rational (or at least quasi-rational) when making
decisions under conditions of uncertainty. For
example, he lists “Methods of Divination” ranging
from the pseudoscientific (tarot cards and tea leaves)
to sophisticated mathematical models.
When we make decisions, we can only do so with
some vision of the future in mind. That mental image
of what the future will look like is a forecast. We
all make forecasts, and we do so every day. From
the mundane “What do I wear today?” to the highly
important “How shall I invest my nest egg?” we
make decisions in the face of unknown, and perhaps
unknowable, information.
It has been said that “all of science is for prediction.”
Although each of us is a forecaster, Orrell reminds us
that forecasting is the scientist’s profession. In The
Future of Everything, he divides “everything” into
three areas of scientific prediction: weather, health,
and wealth. He examines the interconnections among
physical, biological, and economic systems, and
shows how events in one area are inextricably linked
to events in the other two.
For example, a hurricane (think Katrina) can have
huge economic consequences. Economic growth has
environmental consequences (think CO2) that can affect
the weather. An epidemic (think the Black Plague) can
produce economic disruptions with immense impact
that can last for years, even centuries. And obviously
the weather can affect health (think the flu).
Paul is on the faculty of the Statistics Department at Brigham Young University. He is Editor of STATS: The
Magazine for Students of Statistics, a publication of the American Statistical Association, and serves on
Foresight’s Editorial Board.
Summer 2008 Issue 10 FORESIGHT
5
David Orrell, PhD (2007)
THE FUTURE OF EVERYTHING: THE SCIENCE
OF PREDICTION
Thunder’s Mouth Press
ISBN-13: 978-1-56025-975-6
449 pages | $27.95 list price
Orrell maintains that making predictions in one area is
ineffective without considering the interrelationships
with predictions in these other areas. The connections
exist and must be taken into account.
The author quotes Niels Bohr’s famous line, “Prediction
is very difficult, especially if it’s about the future.”
Yet, Orrell wonders, if “weather is just fluid flow, the
human body is biochemistry, the economy is money,”
then why is forecasting so difficult? He notes that
mathematical models should make it easy, but they
don’t. Forecasters should agree, but they don’t. Why
not? Trying to answer that question is what The Future
of Everything is all about.
At the risk of being “Galileo before the Inquisition”
or the little boy who points out that the emperor has
no clothes, Orrell argues that “we need to change our
approach to prediction.” Forecasting methods are
founded on the concept that the past can be used to
predict the future, i.e., the past portends the future.
Orrell raises serious questions about the validity of
such an assumption, and argues that this is a shaky
Table 1. Excerpts from David Orrell’s Predictions for the Year 2100
Most plausible:
• The average global temperature will rise by about five degrees (C or F).
• Storms and hurricanes will dramatically intensify.
• Biotechnologists will accidentally or deliberately create novel pathogens that will be released
into the population.
• The gap between the rich and poor will accelerate, leading to increased social and economic
instabilities.
• The NASDAQ stock index will reach one million.
• Wars will erupt over water, as well as oil.
• There will be a nuclear war, followed by a nuclear winter.
• Civilization will collapse globally.
Other things to look out for:
• The average global temperature will be little changed.
• Partly as a result of a booming global economy, birth rates will fall more quickly than
anticipated. The earth’s population will not be much greater than today’s.
• Most cancers will be curable or treatable in rich countries.
• The NASDAQ stock index will cease to exist.
• Civilization will prosper globally.
Finally, we might:
• Begin to see the planet as a living system, and as a result stop damaging it.
• Denounce our oracles as false, deluding, and distracting – or simply stop listening to them.
6
FORESIGHT Issue 10 Summer 2008
Forecasters should agree,
but they don’t. Why not?
Trying to answer that question
is what The Future of Everything
is all about.
minded me of the importance of such sage forecasting advice as:
• A model is only as good as its underlying
assumptions.
• Avoid overreliance on a single method.
• Constantly collect new information and update
the forecast.
• View all predictions with a dose of critical
skepticism.
Repeatedly, Orrell shows that naïve reliance on
experts’ forecasts might be misplaced confidence,
when looking at their track record. How then can we
trust predictions of the future? I remember a line from
a Bob Dylan song: “You don’t need a weatherman to
know which way the wind blows.” So perhaps the best
forecasting advice is to “make your own forecasts.”
philosophical foundation for a process as fundamental
to decision making as forecasting. He even quotes a
footnote that usually appears on a typical mutual fund
prospectus, “Past performance is no guarantee of
future results,” with the implication that in the spirit of
full disclosure perhaps the same disclaimer should be
a footnote on all forecasts.
The author contends that “Life is not a predictable
machine. Life is a surprise.” He maintains that the
future is unpredictable due to the nonlinear nature of
the complex systems that make up the world. Still, it is
a basic component of human nature to desire a guide
for navigating the uncertain waters of the future. It’s
also clear that models are indispensable to helping us
understand the present world, and to making decisions
about the world we would like to have. Orrell suggests
that even if we cannot predict the future with certainty,
our purpose should be to predict our ability to respond
to the things that could happen.
Orrell concludes The Future of Everything with a list
of what are, in his opinion, the most plausible forecasts
for the year 2100. It is a fascinating list, as you can see
in Table 1. But since predictions made in the past have
been so far off the mark (see “Great Predictions from
History,” beginning on page 343), what makes this list
any more plausible than any other list? It makes me
wonder if a contrarian view might be more reliable
when presented with such a list. Or perhaps his list is
yet another reminder that we should use “the science of
prediction” not to make passive forecasts of the future,
but as a guide to proactively create the future.
The Future of Everything is informative and
entertaining. More importantly, it can change the way
you view forecasting. As Orrell states, his purpose in
The Future of Everything is to “make a forecast about
forecasting.” Read it, and see if your forecast about
forecasting agrees with his.
CONTACT
The Future of Everything helps keep the “science
of prediction” in proper perspective. Reading it re-
Paul J. Fields
Brigham Young University
[email protected]
Summer 2008 Issue 10 FORESIGHT
7
THE INTERNATIONAL JOURNAL OF APPLIED FORECASTING
FORESIGHT melds
practitioner expertise
and cutting-edge
academic research to
bring you forecasting
best practices you can
put to use today.
You'll benefit from
case studies, timely
special features, and
reviews of forecasting
tools and books — all
designed to deepen
your forecasting
knowledge and
advance your
professional status.
Each issue contains authoritative articles
written by top academics and forecasting
practitioners on:
• Forecasting process design and
management
• Forecasting
principles and
methods
• Measurement and
tracking of forecast
accuracy
• Sales and
operations planning
• Forecasting
intelligence, hot new research, and more.
Subscribers receive one year (4 issues) of
FORESIGHT in print, access to our full library
of back issues online, and a subscription to our
quarterly e-Newsletter The Oracle.
Three Easy Ways to Subscribe
$95 for a single subscription
$395 for a Site License (5 print+online)
ONLINE at www.forecasters.org/foresight/subscribe.html
(Please enter docstore in the "Referred By" field)
EMAIL [email protected], or PHONE +1 781 234 4077
FORESIGHT is a publication of the International Institute of Forecasters
5/2010
IIF Business Office | 53 Tesla Avenue | Medford, MA 02155 USA | +1 781 234 4077 | www.forecasters.org/foresight